中文摘要：

分析城市人口流动行为有助于合理分配社会资源，有效应对交通压力、维护社会公共治安等．传统的人工分析方法，如问卷调查、座谈访问等，成本高昂且低效率．智能手机的不断发展与普及在为人们日常生活带来极大便利的同时，所产生的用户移动轨迹数据为有效分析城市人口流动行为提供了可能．然而，海量、低质的轨迹数据给查询分析工作带来了诸多挑战．文中提出了一个分布式人口流动分析框架，采用多节点处理任务，从而提升了算法的执行能力和可扩展性．利用手机运营商提供的手机轨迹数据，分析城市人口流动情况，建立了多个模型，包括进出城市的人口流动行为分析模型、市内各区县间的人口流动行为分析模型、居民工作地／居住地人口分析模型．与传统方法相比，本方案的成本更低，效率更高，覆盖人群更广．

英文摘要：

Analysis on urban population flow can help to make rational distribution of social resources, cope with traffic pressure and maintain public order, etc. The traditional manual analysis methods, such as questionnaire and interview, can not deal with this task efficiently. The continuous development and prevalence of smart phones bring great convenience to people＇s daily life and users＇ trajectory data generated by the connection between smart phones and base stations, which makes it possible to implement this task. However, trajectory data is massive and has low quality, which brings great challenge to related work. We propose a distributed framework for population flow analysis by using multiple computing nodes, thus greatly enhancing efficiency and scalability. In this paper, we use the massive trajectory data to analyze the behavior of urban population flow. We model flowing behavior among cities and among inner-city districts, and decide the work place and living place of each person. Compared with the traditional methods, our method is cheaper and more efficient.